7 November 2022. Food | AI

How to feed the planet. // AI and the corporate capture of the imagination.

Nov 07, 2022

Welcome to Just Two Things, which I try to publish three days a week. Some links may also appear on my blog from time to time. Links to the main articles are in cross-heads as well as the story. A reminder that if you don’t see Just Two Things in your inbox, it might have been routed to your spam filter. Comments are open.

1: How to feed the planet

The climate change writer and activist Jeremy Williams had to give a talk recently on world food to a student conference in China (remotely, I suspect) and has posted a summary to his blog across three posts. The first of the three is here.

His starting point—from George Monbiot’s book Regenesis—is that as a species we’ll need to produce more food on less land, and the talk works through the logic of this.

Only around 20% of the planet’s surface is “useful land” (we have to allow for ice and deserts), and only about half of that is currently used for agriculture—after allowing for lakes, forests, and small amount of useful land (around 1%) that is occupied by cities.

(Source: Jeremy Williams, Earthbound Report)

So in summary, about 10% of the planet’s surface is used for agriculture. At the moment, more than three-quarters of that goes on raising animals and feeding them, and the rest is used for crops.

But: this is all getting worse:

We’re getting all our food from a limited amount of space, but that space is shrinking because of climate change, soil erosion, and declining fertility. At the same time, the number of people who need to be fed is increasing as the global population grows. That means that the amount of farmland per person is going down each year.

Quite dramatically, as well, to judge from a chart in his presentation:

(Source: Jeremy Williams, Earthbound Report)

So it you need to grow more food, you can either increase the amount of land overall, or you can grow more food on each hectare.

These are both problematic—increasing the amount of land basically comes at the expense of forest, which we need to absorb carbon. Growingmore food, so far, has involved using more fertiliser, which has created a vicious circle which leads to soil depletion.

As it turns out, there is a quick win when it comes to getting more food from the same number of hectares, and it’s not about production but about consumption:

Globally, a third of all food is wasted. 1.3 billion tonnes of food is thrown away every year. That’s food that could be feeding people or animals, and so it’s an obvious and practical solution. We’d also save money and reduce greenhouse gases from waste, so everybody wins.

The reasons for these losses are different in the global South and North. In the South, it’s about ‘food loss’ on the way to market, for example because of poor storage.

There are some simple things that can be done about this—crates, for example, reduce the amount of spoiled food during transport. Refrigeration also helps, and he mentions a company in India, Ecofrost, that is making solar powered refrigerators.

In the North, it’s more a problem of food waste. Education helps, and so does regulation. (France has just introduced a law that makes it illegal to throw away edible food.) We’re seeing innovative approaches to using food that would otherwise be wasted—I liked the company that takes leftover cooking water from a hummus factory and uses it to make vegan mayonnaise.

Food waste can also be used to make other types of food. Ynsect, in France, takes food waste, feeds it to insects, and then uses the insects to make animal feed, which should reduce the amount of land used to feed animals.

Of course, this is to some extent fiddling around at the edges. Changing diet away from animals means less land is given over to both growing animals and also feeding them:

A third of the world’s grain is fed to animals. So is a quarter of all the caught fish, because fishmeal is an important protein source in lots of animal feeds. According to a study earlier this year, the food fed to animals could feed another billion people... It takes 16 kilos of grain to produce 1 kilo of beef. If people eat the grain themselves, you can feed more people with the same amount of food.

There are also ways to increase the amount of land you’re growing food on, without deforestation.

One is growing food within the forest—agroforestry, as it is known. Both coffee and cocoa can both be grown under trees. Apparently the coffee realises higher prices because it tastes excellent.

You can grow cows in forests too. Although we’re used to seeing them in fields, they like woods too:

This is called silvopasture and it was very common in Europe and Asia in the past. It’s kind of a forgotten form of farming, but people are bringing it back. There’s a famous example in England at the Knepp estate, where the cows basically roam wild across the woodlands. This is less ‘efficient’ than feedlot beef and more expensive, so we’d still be looking at demand reduction, but it would be better for wildlife and for climate change.

It also seems possible to grow some food in deserts, using some combination of solar energy and a bit of desalination. Chinese researchers have noticed that solar farms in deserts start to reverse desertification—they create a micro-climate, even if these are likely to be small wins in the scale of things.

The oceans are a much bigger win. Jeremy is particularly excited about seaweed, which is a vastly versatile crop that we could eat, or feed to animals, or use for fertiliser, or to make other materials:

(T)here are many benefits to seaweed farming. It frees up land. It doesn’t need any fertiliser. It’s good for marine biodiversity. It locks up carbon... In part 2 I wrote about eating less meat to free up the land that’s used to grow animal feed. What if we grew the animal feed in the oceans? Researchers have found that feeding certain kinds of seaweed to cows can reduce their methane emissions, which are a major greenhouse gas problem.

Along the way, he mentions vertical farming and acquaponics, which are ways of increasing food production in cities and/or in water, depending how you do it.

He ends with what he describes as the “most science fiction sustainable solution I’ve ever come across”: food balls made from combining air and sunlight. Solar Foods is a Finnish firm:

(S)cientists have discovered a micro-organism that feeds on electricity. They feed it solar power, and it reproduces in a big steel vat that’s not very different from the brewing processes used to make yeast. Then you drain it and get a powder that’s almost pure protein, which can be used as an ingredient in bread or pasta or all sorts of other foods.

(Photo: Solein.com, via Earthbound Report)

Who knows if that is going to scale. But reading his three pieces (one, two, and three) I came away more optimistic about our prospects for feeding the planet.

2: AI and the corporate capture of the imagination

There’s lots of noise at the moment about how AI is suddenly becoming a force for creativity. At The Bulletin Annie Dorson suggests that the kind of creativity we’re seeing in the current programs—Dall-E and so on—is a form of corporate plug-and-play.

She’s not against artists using computers in their work. She’s done it herself, as she explains:

In my theater works to date—a combination of algorithmic art and performance—I’ve collaborated with programmers to design and write the code that generates each evening’s show. For an upcoming piece, however, I wanted to explore text- and image-generating software like OpenAI’s GPT-3, Dall-E, and Midjourney. The results were as suspected. Writing one’s own code feels like any other art-making process: full of uncertainty and trial and error, as the artist gropes around for a method that will produce the desired results.

In contrast, using the prompts provided by AL learning models seems like, in her words, ”playing the slots”. One of the problems is that it’s hard to know how changing the inputs will influence the outputs, since—of course—the algorithms are buried deep inside the program, and certainly aren’t available for inspection, or tweaking.

all algorithms have computational bias. They allow certain possibilities and preclude others. This bias is often hidden to the user. If users don’t know why the model responds as it does, or what data it’s using, or how adjusting the code might affect the outcomes, the only real pleasure is in seeing the result. And then seeing what they get if they try it again... Pay another nickel and pull the lever one more time.

She refers specifically to the AI art program Midjourney, although this seems to be down to the fact that its founder, David Holz, was quoted in a somewhat uncritical New York Magazine article recently. Apparently the Midjourney aesthetic is that it likes to “use teal and orange.”

There are obvious concerns here, starting with the cheesiness of the images. Dorson runs through the familiar ones quickly, notably that all of these models have been trained on images scraped off the internet without respecting their creators’ copyrights, while burying the code in a mass of legal and other protections.

However, she has a deeper concern:

These tools represent the complete corporate capture of the imagination, that most private and unpredictable part of the human mind... When tinkerers and hobbyists, doodlers and scribblers—not to mention kids just starting to perceive and explore the world—have this kind of instant gratification at their disposal, their curiosity is hijacked and extracted. For all the surrealism of these tools’ outputs, there’s a banal uniformity to the results.

Drop down menus, in other words, are not a substitute for experimentation and imagination.

And here she uses the word “immiserisation”, which I am more familiar with in the context of the economics of exploitation, but she is borrowing from a philosopher, Bernard Stiegler, who used to write of “symbolic misery”. He used it to describe a world in which our lives are packaged up and then sold to us by corporate interests. (Maybe the ghost of Mark Fisher can be seen between these lines as well).

That’s bad not just because of the dulling sameness of a world of infinite but meaningless variety (in shades of teal and orange). It’s bad because a person who lives in the malaise of symbolic misery is, like political philosopher Hannah Arendt’s lonely subject who has forgotten how to think, incapable of forming an inner life. Loneliness, Arendt writes, feels like “not belonging to the world at all, which is among the most radical and desperate experiences of man.” Art should be a bulwark against that loneliness, nourishing and cultivating our connections to each other and to ourselves.

j2t#390

If you are enjoying Just Two Things, please do send it on to a friend or colleague.